SCENARIOS FOR THE 21

ST

CENTURY

APPENDICES 1. Clines Growth Projections: Closing the Income Gap in 650 Years 2. Model A: Estimation of Initial Values and Rates of Change 3. The Model A Income Distribution Equations 4. Economic Inequality 5. Definitions of Growth, Well-Being and Related Key Concepts 6. Estimates of Stocks of Energy Resources 7. Calculation of Global Warming Trajectories

Appendices 1

SCENARIOS FOR THE 21

ST

CENTURY

APPENDIX 1. CLINES GROWTH RATE PROJECTIONS: CLOSING THE INCOME GAP IN 650 YEARS In 1992 William Cline prepared an influential study on the economics of global warming. In order to estimate benefits and costs of global warming, and of efforts to reduce it, some baseline projections of economic growth over the period of concern are necessary. Clines base-line projections for the more developed countries (MDCs) and the less developed countries (LDCs) are shown in A1-1. In the case of global warming the period of concern is two or three centuries. Equity considerations were not a major focus of Clines study, but that makes a look at the equity implications of his assumed growth trajectories all the more telling. We see that Cline projects that the ratio between the per capita incomes of the more and less developed countries will decline from about 4.3:1 to about 3:1 over the next 275 years. If the growth rates projected for 2275 remain unchanged after that date, the LDCs will catch up with the MDCs after an additional 365 years.1 Per capita incomes in 2650 would be about $390,000, and average income for a family of four would be about $1,600,000.

Given 185,397e^(.002t) = 61,376e^(.005t), then t = 365.

Appendices 2

SCENARIOS FOR THE 21

ST

CENTURY

BOX A1-1. CLINE'S GROWTH RATE PROJECTIONS: CLOSING THE INCOME GAP IN 650 YEARS Source: Cline (1992) per capita income (1990 US $) 2000 MDC's 2025 2050 2075 2100 2125 2150 2175 2200 2225 2250 2275 17,020 22,934 30,902 41,639 56,106 71,953 92,274 112,614 137,438 155,689 176,364 185,397
1.2 1.2 1.2 1.2 1.2 1.0 1.0 0.8 0.8 0.5 0.5 0.2

growth rate (%/yr)

LDC's
growth rate (%/yr)

3,985
1.5

5,782
1.5

8,389 12,172 15,610 20,019 25,673

1.5 1.5 1.0 1.0 1.0

32,924
1.0

42,222
1.0

47,829
0.5

54,181
0.5

61,376
0.5

MDC's/LDC's

4.27

3.97

3.68

3.42

3.59

3.59

3.59

3.42

3.26

3.26

3.26

3.02

FIGURE 1.

RATIOS OF PER CAPITA INCOMES OF MDC'S AND LDC'S (source: Cline 1992) 4.50 4.00 3.50 3.00
ratios

2.50 2.00 1.50 1.00 0.50 0.00 2000 2025 2050 2075 2100 2125 2150 2175 2200 2225 2250 2275

TABLE 1.

COMPARISONS OF PER CAPITA INCOMES PROJECTED BY CLINE Per Capita Income (1990 US $) World 1995 2100 2275
6,319 21,000 77,700 "More developed countries"

family of 4
25,276 84,000 310,000

output growth rate

2275 185,400
"Less developed countries"

742,000

.002/year

2275

61,400

246,000

.005/year

Appendices 3

SCENARIOS FOR THE 21

ST

CENTURY

APPENDIX 2.

MODEL A: ESTIMATION OF INITIAL VALUES AND RATES OF CHANGE

For the Reference Scenario (Scenario I) I needed to specify initial values for population, per capita GDP, distribution of income by quintile, and energy intensity for each of the four world income/geographic sectors, i.e., high income, middle income, China and low income. For each of these variables and sectors I also needed to estimate an initial value for the rates at which each are changing. I also needed to specify rates of change during successive periods over the 150 years covered in the reference scenario of this model. The general procedure I used was to review several to many existing studies and use these to choose sets of values that appeared to have broad support and that I believed to be reasonable otherwise. Rates of change for population, per capita GDP and income distribution used in the five policy scenarios (Scenarios 2 through 6) are motivated in the main text. The process used to estimate rates of change of energy intensity used in the five policy scenarios is described in this appendix. Sources for all estimates are shown in Box A2-1, and the values projected by these sources are shown in the additional boxes.

I. INITIAL VALUES For initial values I relied on those shown in the 1996 World Development Report. The only modification was to separate China from the Reports low income category and show it as a separate sector. Calculations for population and per capita GDP are shown in A2-2. Estimates of energy intensity are shown in A2-3. The most accurate way to compute the share of output received by income quintiles in a sector is to calculate mean incomes for each quintile of each country in that sector and then aggregate these, from lowest to highest, into new quintiles. I did this for the high income sector but the difference between using this technique and simply taking the mean of the values given

Appendices 4

SCENARIOS FOR THE 21

ST

CENTURY

BOX A2-1. Sources Consulted for Population, GDP and Energy Growth Rate Projections
1. Energy Modeling Forum-14. 1996. John Weyant, co-ordinator. Stanford University. 2. Holdren, John. 1995. World Energy Futures, class handout for ER 100/200, University of California, Berkeley. 3. Intergovernmental Panel on Climate Change. 1994. Climate Change 1994: Radiative Forcing of Climate Change and an Evaluation of the IPCC IS92 Emission Scenarios. 4. Lazarus, Michael., and Lisa Greber, Jeff Hall, Carlton Bartels, Steve Bernow, Evan Hansen, Paul Raskin, and David Von Hippel (1993). Towards a Fossil Free Energy Future: The Next Energy Transition. A Technical Analysis for Greenpeace International. Boston, Stockholm Environment Institute - Boston Center; Tellus institute. 5. Londono, Juan Luis. 1996. Poverty, Inequality, and Human Capital Development in Latin America, 1950-2025. Washington, D.C.: The World Bank. 6. Manne, Alan S., Robert Mendelsohn and Richard Richels. 1993. MERGE - A Model for Evaluating Regional and Global Effects of GHG Reduction Policies. Stanford University, Yale University and Electric Power Research Institute. 7. Sachs, Jeffrey, Growth in Africa: It Can Be Done. The Economist, June 29, 1996. 8. Tol, Richard S. 1993. The Climate FUND. Vrije Universiteit. Amsterdam, Holland. 9. World Bank. 1996. World Development Report. 10. World Energy Council/IIASA. 1995. Global Energy Perspectives to 2050 and Beyond. Laxemburg, Austria.

Appendices 5

SCENARIOS FOR THE 21

ST

CENTURY

BOX A2-2. INITIAL VALUES FOR POPULATION, GDP, AND AVERAGE ANNUAL GROWTH RATES, 1990-94 source: World Development Report, 1996. The World Bank Calculated values that depend on growth rates use the exponential formula, r = ln(Q1/Q2)/n. A. SEPARATION OF CHINA from LOW INCOME figures GDP: GDP '94 x10^12 (tbl 12, p210) 1.21 0.52 growth rate, 90-94 (%/yr) (tbl 11, p208) 6.20 12.90 GDP '90 x10^12 (calculated) 0.94 0.31 POP 90 x10^6 (calculated) 2960.95 1135.18 GDP/pop 90 318.47 274.57 1990

LOW INCOME COUNTRIES (including China) CHINA POPULATION:

POP 94 growth rate, 90-94 x10^6 (%/yr) (tbl 4, p 194-5) ==> 3182.00 1191.00 1.80 1.20

LOW INCOME (including China) CHINA Per Capita GDP: LOW INCOME (including China) CHINA

GDP/pop 94 growth rate, 90-94 (calculated) ==> 379.76 438.43 4.40 11.70

1994 growth rate, 90-94 (calculated) LOW INCOME COUNTRIES (WITHOUT CHINA): GDP POPULATION GDP/POP (x10^12) (x10^6) ($) 0.69 1991.00 344.66 2.09 2.17 -0.08

0.63 1825.77 345.77 1990 4037.57 1478.57 2730.73

B. MIDDLE INCOME COUNTRIES: GDP POPULATION GDP/POP (x10^12) (x10^6) ($)

1994 growth rate, 90-94 (average, %/yr) 4070.00 1570.00 2592.36 0.20 1.50 -1.30

C. HIGH INCOME COUNTRIES: GDP POPULATION GDP/POP (x10^12) (x10^6) ($) 20120.00 850.00 23670.59 1.70 0.70 1.00 18797.32 826.53 22742.45

sources for middle and high income countries: population '94 and population growth rate, '90-'94: Table 4 p194 GDP growth rate, '90-'94: Table 11 p 208 GDP '94: Table 12 p 210 all other values are derived

Appendices 6

SCENARIOS FOR THE 21

ST

CENTURY

BOX A2-3. INITIAL VALUES FOR ENERGY INTENSITY source: World Development Report, 1996. Tables 8 and 12 (kg = kg of oil equivalent} (1 kg oil/year = 1.42 W) total commercial energy use, 1994 (x 10^6 kg) (TW) all low income China low income, minus China middle income high income 1222928 770000 1.737 1.093 GDP, 1994 (x10^6) ('87 US $) 1208422 522172 GDP/kg ('87 US $) Watts/$GDP ('87 US $)

0.678

2.094

452928 2501145 4392058

0.643 3.552 6.237

686250 4069532 20120240

1.515 1.627 4.581

0.937 0.873 0.310

TOTAL in WDR:

8116131 8035058

11.525 11.410

25398194 25223462

3.129 3.139

0.454 0.452

Appendices 7

SCENARIOS FOR THE 21

ST

CENTURY

for each country was small. For the remaining three sectors I simply used the mean of the quintile shares for each country in a sector, as shown in A2-4. The values shown for each quintile incorporate the simplifying assumption that the entire GDP is received by households. I dont believe that adjustment for this tendency would greatly affect the overall results of the scenario exercises, but extended treatments of the topics addressed in this exercise might test this assumption.

II. GROWTH RATE PROJECTIONS I used the studies listed in Box A2-1 to develop assumptions concerning the rates of change of population, per capita GDP and energy intensity. A. Population I used the population growth rate assumptions used by the IPCC and Tellus, adjusted to account for the different aggregation of countries and regions. B. Per Capita GDP Summaries of the assumptions concerning growth of per capita GDP used in the source studies are shown in A2-5. 1. Near-term per capita GDP growth rates (1990-2000) A. low income countries The 1996 World Development Report shows negative growth for 1990-94 among the low income countries. However, the greatest part of this reflects the difficulties of 3 newly independent states (Tajikistan, Azerbaijan and Armenia), and Rwanda. The largest economies saw positive growth (India-2.9%; Pakistan-1.3%, Bangladesh-2%, Nigeria-1.2%). Its reasonable to assume that the aggregate negative growth results are temporary. Thus I chose 1% for the sector for the period 1994-2000. This is not inconsistent with the figures shown by other studies. B. China

Appendices 8

SCENARIOS FOR THE 21

ST

CENTURY

BOX A2-4. INITIAL VALUES FOR INCOME BY POPULATION QUINTILES source: World Development Report, 1996. The World Bank. A. LOW INCOME COUNTRIES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 Rwanda Mozabique Ethiopia Tanzania Burundi Sierra leone Malawi Chad Uganda Madagascar Nepal Vietnam Bangladesh Haiti Niger Guinea-Bissau Kenya Mali Nigeria Yemen,Rep. Burkina Faso Mongolia India Lao PDR Togo Gambia, The Nicaragua Zambia Tajikistan Benin Cent. Af. Rep Albania Ghana Pakistan Mauritania Azerbaijan Zimbabwe Guinea China (see below) Honduras Senegal Cote d'Ivoire Congo Kyrgyz Rep Sri Lanka Armenia Cameroon Egypt Lesotho Georgia Myanmar Population (10^6) 1991 GDP (10^6) 686270 6.16 106.12 10.18 175.52 14.55 250.69 21.10 363.71 48.01 827.45 = 100 Q1 9.7 Q2 13.2 Q3 16.5 Q4 21.6 Q5 39.1

6.9

10.9

15.3

21.5

45.4

6.8 5.8 9.1 7.8 9.4 7.5 2.1 3.4 4

10.3 9.9 12.9 11.4 13.5 11.8 6.5 6.7 8.9

14.4 14 16.7 15.4 17.2 15.5 12 10.7 14.4

20.4 20.3 21.8 21.4 22 21.1 20.6 17 23.4

48.1 50 39.5 44 37.9 44.1 58.9 62.1 49.3

8.5 9.6

12.1 12.9

15.8 16.3

21.1 21

42.6 40.2

4.2 3.9

8 8

12.6 13.8

20 23.8

55.2 50.4

7.9 8.4 3.6 4 3 3.8 3.5 6.8

12 12.9 10.6 6.3 8.3 7.4 7 11.2

16.1 16.9 16.2 10 14.6 12 11.6 15.8

21.8 22.2 23 17.4 23.9 19.4 19.3 22.2

42.2 39.7 46.5 62.3 50.2 57.4 58.6 44.1

8.9

13.1

16.9

21.7

39.3

8.7 2.8

12.5 6.5

16.3 11.2

21.4 19.4

41.1 60.1

totals:

Mean share of income for the Quintile (%) Mean Income for the Quintile ($): B. CHINA totals: Population (10^6) 1191 GDP (10^6) 522172

Q1 6.20 135.91

Q2 10.50 230.18

Q3 15.80 346.36

Q4 23.60 517.35

Q5 43.90 962.36 = 100

Mean share of income for the Quintile (%) Mean Income for the Quintile ($):

[more]

Appendices 9

SCENARIOS FOR THE 21

ST

CENTURY

BOX A2-4. Initial Values (cont.) C. MIDDLE INCOME COUNTRIES 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 Bolivia Macedonia Moldova Indonesia Philippines Uzbekistan Morocco Kazakstan Guatemala Papua New Guinea Bulgaria Romania Ecuador Dominican Rep Lithuania El Salvador Jordan Jamaica Paraguay Algeria Colombia Tunisia Ukraine Namibia Peru Belarus Slovak Rep Latvia Costa Rica Poland Thailand Turkey Croatia Panama Russian Fed Venezuela Botswana Estonai Iran Turkemistan Brazil South Africa Maruitius Czech Rep Malaysia Chile Trin & Tob Hungary Gabon Mexico Uruguay Oman slovenia Saudi Arabia Greece Argentina Korea, Rep. Population (10^6) totals: 1570 GDP (10^6) 4069532 6.39 828.09 10.43 1352.30 14.74 1910.96 21.11 2735.30 47.34 6134.99 = 100 [more] Q1 5.6 6.9 8.7 6.5 6.6 7.5 2.1 8.3 9.2 5.4 4.2 8.1 5.9 5.8 6.9 3.6 5.9 9.5 4.9 11.1 11.9 9.6 4 9.3 5.6 Q2 9.7 11.9 12.3 10.1 10.5 12.3 5.8 13 14.4 8.9 7.9 12.3 9.8 10.2 11 7.6 10.4 14.1 9.2 15.3 15.8 13.6 9.1 13.8 8.7 Q3 14.5 16.7 16.3 14.4 15 16.9 10.5 17 18.4 13.2 12.5 16.2 13.9 14.9 15.1 12.6 15.3 18.1 14.1 18.5 18.8 17.5 14.3 17.7 13 Q4 22 23.1 22.1 21.2 21.7 22.9 18.6 22.3 23.2 19.9 19.7 21.3 20.3 21.6 20.9 20.4 22.1 22.9 21.4 22.2 22.2 22.6 21.9 22.6 20 Q5 48.2 41.5 40.7 47.8 46.3 40.4 63 39.3 34.8 52.6 55.7 42.1 50.1 47.5 46.1 55.8 46.3 35.4 50.4 32.9 31.4 36.7 50.7 36.6 52.7

2 3.7 3.6 6.6 6.7 2.1 3.3 10.5 4.6 3.5 9.5 4.1

6.3 8.5 7.1 10.7 11.4 4.9 5.8 13.9 8.3 6.6 14 7.8

11.6 13.5 11.7 15.1 16.3 8.9 9.8 16.9 13 10.9 17.6 12.5

20.3 20.4 19.3 21.4 22.8 16.8 17.7 21.3 20.4 18.1 22.3 20.2

59.8 53.8 58.4 46.3 42.8 67.5 63.3 37.4 53.7 61 36.6 55.3

9.5

13.5

17.1

21.9

37.9

Mean share of income for the Quintile (%) Mean Income for the Quintile ($):

Appendices 10

SCENARIOS FOR THE 21

ST

CENTURY

BOX A2-4. Initial Values (cont.) D. HIGH INCOME COUNTRIES population GDP (10^6) (10^6) 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 Portugal New Zealand Spain Ireland Israel Australia UK Finland Italy Kuwait Canada Hong Kong Netherlands Singapore Belgium France Sweden Austria Germany USA Norway Denmark Japan Switzerland UAE totals: 10 3 39 4 5 18 58 5 57 2 29 6 15 3 10 58 9 8 82 261 4 5 125 7 2 87257 50777 482841 52060 77777 331990 1017306 97961 1024634 24289 542954 131881 329768 68949 227550 1330381 196441 196546 2045991 6648013 109568 146076 4590971 260352 35405 20% of pop (10^6) 2 0.6 7.8 0.8 1 3.6 11.6 1 11.4 0.4 5.8 1.2 3 0.6 2 11.6 1.8 1.6 16.4 52.2 0.8 1 25 1.4 0.4

Q1

Q2

Q3

Q4

Q5

5.1 8.3 6 4.4 4.6 6.3 6.8 5.7 5.4 8.2 5.1 7.9 5.6 8 7 4.7 6.2 5.4 8.7 5.2

10.8 13.7 12.1 11.1 10 12.1 12 11.8 10.8 13.1 9.9 13.7 11.8 13.2 11.8 11 12.8 12 13.2 11.7

16.2 18.1 17.8 17.5 16.8 18.4 16.7 17.7 15.2 18.1 14.6 18.6 17.2 17.4 17.1 17.4 18.9 18.4 17.5 16.4

23.2 23.4 24.5 24.8 24.3 25.5 23.5 24.6 21.6 23.7 21.4 23.8 23.5 24.5 23.9 25 25.3 25.6 23.1 22.1

44.7 36.6 39.6 42.2 44.3 37.6 41 40.2 47 36.9 48.9 36 41.9 36.9 40.3 41.9 36.7 38.6 37.5 44.6

850 20120240 6.23 7373.48 11.93 14119.67 17.30 20475.30 23.87 28245.27 40.67 48134.72 = 100

Mean share of income for the Quintile (%) Mean Income for the Quintile ($):

Appendices 11

SCENARIOS FOR THE 21

ST

CENTURY

BOX A2-5. PER CAPITA GDP GROWTH RATE ASSUMPTIONS This box displays the values that were used in a variety of studies as estimates of the rate at which per capita GDP would grow over the course of several to many decades in various parts of the world. These values were used to set per capita GDP growth rates used in the Model A reference scenario. As described in the Appendix 2 text, I chose values that appeared to have broad support, and that otherwise seemed reasonable. A. LOW INCOME COUNTRIES [All values shown are percent per year] 1. Tellus (based on IPCC 92) Africa: 1990-2030: 2030-2100: Southeast Asia: 1990-2030: 2030-2100: Middle East: 1990-2030: 2030-2100: .90 2.48 2.50 2.90 .90 1.90

2. World Energy Council/IIASA (total growth rate, not per capita growth rate) Africa: 1990-20: hi = 3.3; mid = 3.0 2020-50: hi = 4.7; mid = 3.5 South Asia: 1990-20: hi = 3.9; mid = 3.6 2020-50: hi = 4.6; mid = 3.5 3. Energy Modeling Forum other non-OECD: 1990-2000: 1.44 2000-2025: 2.55 4. Tol (based on Manne-Mendelsohn-Richels) Africa: 1990-00: .0 2000-10: .25 2010-20: .75 2020-30: 1.35 5. IPPC (survey of 21 studies) Africa, 1990-2020: minimum: .01 median: 1.50 average: 1.56 max: 2.98 6. Sachs, The Economist, 6/29/96 a. Africa: between 87-94, output per capita fell at a rate of 0.6% per year. The IMF and World Bank project 1-2 percent annual growth rates for the coming decade or so. Sachs calculates that the realistic goal should be 4.4%, which is only slightly (.5%) below the rate for other developing countries. b. Asia: output grew by 7% in 1995. [more ]

Appendices 12

SCENARIOS FOR THE 21

ST

CENTURY

BOX A2-5. Per Capita GDP Growth Rate Assumptions (cont.) B. CHINA 1. Tellus (following IPCC 92) Centrally Planned Asia [CPA]: 1990-2030: 3.1 2030-2100: 3.1 2. World Energy Council/IIASA (total growth rate, not per capita growth rate) CPA: 1990-2020: hi = 7.2; mid = 5.0 2020-2050: hi = 4.4; mid = 4.0 3. Energy Modeling Forum China: 1990-2000: 2.71 2000-2025: 2.66 4. Tol (after Manne-Mendelsohn-Richels) China: 1990-2000: 6.5 2000-2010: 5.0 2010-2020: 3.25 5. IPPC SURVEY (21 studies) CPA, 1990-2020: minimum: 2.01 median: 3.91 mean: 3.73 max: 5.07 [more]

Appendices 13

SCENARIOS FOR THE 21

ST

CENTURY

BOX A2-5. Per Capita GDP Growth Rate Assumptions (cont.) C. MIDDLE INCOME COUNTRIES 1. Tellus (following IPCC 90) Latin America: 1990-2030: 1.9 2030-2100: 2.2 Eastern Europe: 1990-2030: 2.9 2030-2100: 2.0 2. World Energy Council/IIASA (total growth rate, not per capita growth rate) Eastern Europe: 1990-2020: hi = 2.3; mid = 0.9 2020-2050: hi = 4.6; mid = 3.6 3. Energy Modeling Forum Former Soviet Union: 1990-2000: -2.0 2000-2025: 3.9 4. Tol (after Manne-Mendelsohn-Richels) Soviet Union &Eastern Europe: 1990-2000: 2.0 2000-2010: 1.75 2010-2020: 1.9 Tol (cont.) Latin America: 1990-2000: 2000-2010: 2010-2020: 2020-2030: 1.0 1.0 1.25 1.25

5. IPCC SURVEY (21 studies): Eastern Europe & FSU, 1990-2020: minimum: 0.77 median: 1.73 mean: 1.85 max: 3.8 6. Londono Latin America per capita GDP grew at 1.9% in first part of 90s, and could maintain this growth for the next 10 years. [more]

Appendices 14

SCENARIOS FOR THE 21

ST

CENTURY

BOX A2-5. Per Capita GDP Growth Rate Assumptions (cont.) D. HIGH INCOME COUNTRIES 1. Tellus (following 1PCC 92) USA 1990-2030: 2030-2100: Western Europe 1990-2030: 2030-2100: 1.6 1.3 1.9 1.3

2. World Energy Council/IIASA (total growth rate, not per capita growth rate) North America: 1990-2020: hi = 2.3; mod = 2.0 2020-2050: hi = 1.6; mod = 1.4 Western Europe: 1990-2020: hi = 2.2; mod = 1.9 2020-2050: hi = 1.6; mod = 1.3 3. Energy Modeling Forum USA: 1990-2000: 2.18 2000-2025: 1.97 EEC: 1990-2000: 2.15 2000-2025: 2.09 4. Tol (Based on Manne-Mendelsohn-Richels) OECD-America: 1990-2010: 1.5 2010-2020: 1.75 5. IPCC SURVEY (21 studies) USA, 1990-2020: minimum: .90 median: 1.96 mean: 1.91 max: 2.91 6. Londono: per capita growth rate projected for industrialized countries is 2.4% for next decade.

E. TOTAL WORLD: 1. Greenpeace, following IPCC 90 total: 1990-2030: 1.3%; 1985-2100: 1.6% 2. Londono: world, 1991-1994: 1.9%

Appendices 15

SCENARIOS FOR THE 21

ST

CENTURY

The studies listed differ greatly in the assumptions they make about Chinas near-term economic growth. The World Development Report shows 11.7% for 1990-94. The Energy Modeling Forum estimates, which use an average of PPP and exchange rates, shows 2.71% for 1990-2000. The other studies show estimates of 3, 4, 5, and 6.5%. For the time being I chose the middle-range value of 5% as the growth of per capita income in China over the period 19942000. C. Middle Income Countries The recent negative growth rates shown by the World Bank for the middle income countries again largely reflect the experience of the newly independent states. The other countries in this sector are growing. In the other studies Latin America is projected to grow at about 1-2 %. I chose a figure of 1% as a reasonable expected growth rate of middle income countries between 1994 and the end of the decade. D. High Income Countries All the studies showed per capita GDP growth rates of 1.5-2% for the high income countries for 1990-2000. Since the historical rate is 1% for 1990-94, I chose 2% for 1994-2000.

II .Reference Scenario - Long Term Per Capita GDP Growth Rates (2000 - 2150) I used the IPCC figures as reported by Tellus and IPCC as the basis for my per capita GDP growth rate estimates for 2000-2100. I used the values given in EMF-14 to help estimate these rates for 2100-2150. The ten IPCC regions can be assigned to my four sectors roughly as follows: Low Income: African + Southeast Asia China: Centrally Planned Asia Middle Income: Latin America + Middle East + USSR + Eastern Europe High Income: JANZ + US + Western Europe The Tellus per capita GDP growth rate assumptions give these results when recalculated using the 4-sector aggregation:

Appendices 16

SCENARIOS FOR THE 21

ST

CENTURY

1. Low 2. China 3. Middle 4. High

2000-2030 1.9 3.1 1.9 1.8

2030-2100 2.7 3.1 1.9 1.3

2100-2150 2.0 2.0 1.2 .8

2150-2200 1.0 1.0 1.0 .6

I adjusted these rates to show smoother transitions between decades, without changing the final output levels in a large way.

C. Energy Intensity I used five studies listed in Box A2-1 to choose projected values for changes in energy intensity over time. Summaries of the values used in each of these studies are shown in A2-6. I based the estimate for the reference scenario, Scenario 1, on those used by Manne-MendelsohnRichels, and by Tol. The rates used in Scenario 2 are based on Holdrens estimates, but with an imposed upper limit of 2% on the annual rate of energy intensity improvement in the low and middle income sectors, and China. For Scenarios 3 to 6 I used Holdrens full estimates, including the very high rates of efficiency improvement he shows for the non-industrial world in the middle part of the coming century. I slowed the rates down in the final three decades before 2150 to offset these very high, continuing rates of improvement just a bit. The three scenarios are displayed in A2-7. All values shown are annual percentage declines in the ratio of energy use to GDP.

Appendices 17

SCENARIOS FOR THE 21

ST

CENTURY

BOX A2-6. ENERGY INTENSITY CHANGE ASSUMPTIONS

Assumptions concerning the rate of change in energy intensity over coming decades are shown for five studies. These assumptions were used to help specify rates of change used in the Model A reference scenario and the several policy scenarios, as shown in Box A2-7. Citations are shown in Box A2-1. Energy intensity is the ratio of energy use to GDP. It is the inverse of energy efficiency, the ratio of GDP to energy use. A decline in energy intensity is an increase in energy efficiency. 1. Intergovernmental Panel on Climate Change (IPCC) Values chosen for the IPCC scenarios, based on a review of 21 studies covering 1990-2020. World CHINA & CPA 1. NO INTERVENTION ( = business-as-usual) max 1.84 2.85 med .94 1.59 mean .99 1.52 min .53 .22 2. POLICY max med mean min 2.4 1.78 1.74 1.08 4.32 1.89 2.24 .53 AFRICA .67 .17 .30 .53 2.19 1.46 1.31 .30 USA 2.17 1.29 1.35 0.20 2.88 1.91 1.87 .66

Calculation of overall energy intensity decline, 1900-2100: USA .904 CPA 1.46 EU & FSU 1.17 AFRICA .543 2. Energy Modeling Forum 14 (1996) Region USA EEC other OECD CHINA FSU other non-OECD 1990- 20002000 2025 .70 .63 .70 .67 .72 .67 .87 .85 -.66 1.25 .46 .81 date 20252050 .52 .51 .51 .94 1.07 .77 20502075 .37 .37 .36 .92 .60 .76 20752100 .36 .36 .36 .91 .62 .83 21002150 .26 .26 .26 .64 .32 .64 21502200 .19 .19 .19 .32 .26 .32

EMF calculations are based on a formula by Arnulf Grubler of IIASA. Aggregate Energy Efficiency Improvements (AEEI) are a constant fraction, .32, of GDP/capita growth rates. Thus, if USA per capita growth rate for 1990-2000 is 2.18, short run AEEI is .32 x 2.18 = .70, as shown. [more.]

Appendices 18

SCENARIOS FOR THE 21

ST

CENTURY

Box A2-6. Energy Intensity Change Assumptions (cont.) 3. Mann-Mendlesohn-Richels (1993); also used by Tol (1993) OECD ME, LA, AFR, SA, SEA CHINA FSU & EE 4. HOLDREN (1996) 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 Reference case: Industrialized 1.3 1.3 1.2 1.2 1.1 1.1 1.0 1.0 .90 .90 .80 Developing 0 .5 1.0 1.3 1.3 1.2 1.2 1.1 1.1 1.0 1.0 Maximum feasible improvements: Industrialized 1.5 2.5 3.0 3.0 2.7 2.4 2.0 1.5 1.0 1.0 1.0 Developing .5 1.0 2.0 2.3 2.5 2.7 2.8 2.7 2.4 1.8 1.0 5. World Energy Council/IIASA (1995) High Growth World 1990-2050 1.0 1990-2100 1.0 Regions (1990-2050) OECD 1.2 Reforming Economies 2.1 Developing Countries 1.6 Middle Course 0.7 0.8 1.1 1.7 1.2 Ecologically Driven 1.4 0.5 2.0 2.2 1.9 90-00 .50 0 1.00 .25 00-10 .50 0 1.00 .25 10-20 .50 .10 .90 .30 20-30 .50 .20 .80 .35 30-40 .50 .30 .70 .40 40-50 2050-2100 .50 .50 .40 .50 .60 .50 .45 .50

Appendices 19

SCENARIOS FOR THE 21

ST

CENTURY

Appendices 20

SCENARIOS FOR THE 21

ST

CENTURY

APPENDIX 3. MODEL A INCOME DISTRIBUTION EQUATIONS The equations shown in Boxes IA-10 and IA-11 of the text allow us to model changes in the distribution of income among quintiles over time. Box A3-1, below, illustrates how this works.

Box A3-1. Model A Income Distribution Procedure

y5

y4
X4 D1 D2 D3 X5

y y2 y1 Q1 Q2 Q3 Q4 Q5 y3

At the end of each period proportion r of the total income above the mean income y [r (X4 + X5) ] is redistributed to the three quintiles below the mean in the same proportion s that each of them is below the mean. For example, for the bottom quintile, Q1, this proportion is: s1 = D1/(D1 + D2 + D3). We call s the redistributive proportionality factor. This value can be expressed in terms of the mean income of the quintile (yj) and the mean income of the sector as a whole (y), as: s1 = (y y1 ) / [(y y1)+( y y2)+( y y3)] = (y y1) / [ 3y - (y1 + y2 + y3 ) ] Which is equivalent to equation (6) shown in IIA-11 of the text (without the income sector subscript i). The values of s for the three quintiles to which it applies Q1, Q2 and Q3 are shown for all four income sectors in A3-2.

Appendices 21

SCENARIOS FOR THE 21

ST

CENTURY

BOX A3-2.

MODEL A REDISTRIBUTIVE PROPORTIONALITY FACTOR ("s") 2 3 Quintile Mean Income 4 Sector Mean 344.8 Q1 Q2 Q3 Q4 Q5 106 176 251 364 827 438.2 Q1 Q2 Q3 Q4 Q5 136 230 346 517 962 2592.2 828 1352 1911 2735 6135 23669.6 Q1 Q2 Q3 Q4 Q5 7373 14120 20475 28245 48135 29040.8 0.5612 0.3288 0.1100 3685.6 0.4787 0.3365 0.1848 602.6 0.5015 0.3455 0.1530 5 Total Gap 501.4 0.4763 0.3367 0.1871 6 Value of "s"

1 Income Sector low income

China

middle income Q1 Q2 Q3 Q4 Q5 high income

Notes: Columns 1, 2 and 3 identify the four income sectors, the ve quintiles for each sector, and the per capita income of those each of those quintiles. Column 4 shows the mean (or per capita) income for each income sector as a whole Column 5 shows the "total gap" for each income sector, i.e., the sum of the amounts by which each of the rst three quintiles are below the mean. Column 6 shows the percent of the "total gap" represented by the "gap" of each quintile in a sector. This is the value of the redistributive proportionality factor, "s".

Appendices 22

SCENARIOS FOR THE 21

ST

CENTURY

If a positive r is maintained over time the incomes for all quintiles will converge. A negative r will redistribute income from the three quintiles below the mean to the two above the mean. The model allows a growth rate g to be exogenously specified for the economy as a whole. Thus the model can show how the per capita incomes of any quintile, and of the income sector as a whole, will change given assumptions about the aggregate growth rates of income and rates of change in the distribution of income. The figures in A3-3 illustrate Model As income distribution dynamics. Figure A shows the quintile distribution of income for the United States in 1990, in 1990 dollars, and illustrates the trivial scenario in which there is no change in the level of total income (g = 0) , and no change in the distribution of income (r = 0) , over the period between 1990 and 2100. Figure B illustrates a scenario in which the economy grows at a constant rate of 1% per year (g = .01) but the distribution of income is unchanged (r = 0). Both the 80/20 ratio and the Gini Coefficient remain constant over time. However, the absolute difference between the incomes received by each quintile becomes greater. Figure C shows total income growing at the same rate of 1% as in Figure 1b, but now shows redistributive pressure of 1% (r = .01). This means that 1% of the income of the top two quintiles that is above the mean of all five quintiles is redistributed to the bottom three quintiles, at the end of each period. Now the 80/20 ratio and Gini Coefficients decline over time, while the absolute difference among quintiles remains constant. Figures D and E show the results of progressively stronger redistributive efforts, r = .0175 and r = .04, respectively. For any given rate of aggregate growth (in this example, 1%) there is a threshold level of redistributive effort above which the top quintile will be forced to undergo an absolute reduction in its income, rather than simply a reduction in its relative income share. In Figure D (r = .0175) incomes converge rapidly but at no point need the top income quintile forsake further income growth entirely. In Figure E however (r = .04), the incomes of the

Appendices 23

SCENARIOS FOR THE 21

ST

CENTURY

BOX A3-3. Model A Income Distribution Dynamics

Appendices 24

SCENARIOS FOR THE 21

ST

CENTURY

top quintile must decline until 2025, at which time they can begin to grow once more. Note that as the average incomes of the quintiles converge, their rates of growth do so as well. Figure F illustrates a scenario in which there is no aggregate economic growth (g = 0) but the distribution of income becomes more equal. In this situation there is no choice but that the top quintiles undergo an absolute decline in their average incomes. It is instructive to compare Figure F with Figure C. Both show a redistributive pressure of 1% (r = .01) and thus show the same 80/20 ratios and Gini coefficients at every point in time. But in C this takes place while aggregate income is growing at 1%, while in F there is no growth of aggregate income. Interpretation of r As noted in the text, the term redistributive pressure is used here in a very general sense. It might refer to policies that take from the rich and give directly to the poor. It might refer to policies that tax the rich to provide educational services that increase the earnings abilities of the poor. Or it might refer to market mechanisms that are thought to generate greater income equality. If we think of r as a tax, how heavy a tax does a value of r of say, 1%, imply? Suppose per capita income growth in the United States is expected to average 1% over the coming decades, and that in the absence of redistributive policy this growth would be shared equally by all income quintiles. This is the scenario of Figure B. Now suppose we wished to effect a more equal distribution of incomes, say along the lines shown in Figure C. What proportion of their total incomes would the upper quintiles have to forego, and how big a transfer would this represent for the lower quintiles? We see in Table 1 of A3-4 that a value of r of .01 implies a tax on the top quintile (Q5) of .495%, or $450 on a total household income of $90,900. The 4th quintile (Q4) is taxed at .211%, or $120 on a total household income of $56,820.

Appendices 25

SCENARIOS FOR THE 21

ST

CENTURY

BOX A3-4. MODEL A GROWTH AND REDISTRIBUTION DYNAMICS Table 1 shows that for a single period, an "r" value of .01 represents transfers of $450, or less than 1/2 of 1%, of the income of the fth quintile to the bottom three quintiles. Table 1. Growth and Redistribution over one period (g = .01, r = .01) mean household income before after dollars redistribution redistribution transferred quintile Q5 Q4 Q3 Q2 Q1 90900 56820 38640 27270 13650 90450 56700 38700 27450 13950 -450 -120 60 180 300

tax rate/ grant rate (%) -0.495 -0.211 0.155 0.660 2.198

Table 2 shows that over 60 periods an "r" value of .01 represents a much larger transfer of income. By the nal period over 22% of the income of the fth quintile is being transferred to those in the bottom three quintiles. Table 2. Q5 year 1990 2010 2030 2050 Growth and Redistribution over 60 periods ( g = .01, r = .01) mean household income before after dollars redistribution redistribution transferred 90900 109830 134010 163500 -450 -9360 -22020 -36750

tax rate (%) -0.495 -8.522 -16.432 -22.477

80/20 ratio 6.67 4.29 3.16 2.52

100470 111990 126750

Q1 year 1990 2010 2030 2050

mean household income before after dollars redistribution redistribution transferred 13650 16470 20100 24540 300 6930 15390 25710

grant rate (%) 2.198 42.077 76.567 104.768

80/20 ratio 6.67 4.29 3.16 2.52

23400 35490 50250

Appendices 26

SCENARIOS FOR THE 21

ST

CENTURY

Whether these are considered to be heavy taxes or not depends on the story that the model is interpreted to be telling. If the tax dollars are thought to be used for purposes that will increase the earnings of the lower three quintiles in the subsequent period by the amount of the transfer, then the effective tax rate can stay near the low levels shown in the first period. However, if the model is thought to describe a system of direct income subsidies, then the amounts taxed and transferred become much larger. In this case we assume that the earned incomes of the lower quintiles increase by no more than 1% a year. In order to effect a meaningful redistribution the amounts transferred from the upper quintiles (the incomes of which are also increasing at 1% annually) must increase each year. This is shown in Table 2 in Box A34. A redistributive pressure of 1%, applied continuously, requires the effective tax rate on household incomes in Q5 in the years 2010, 2030 and 2050 to grow to 8.5%, 16%, and 22%, respectively. Similarly, by 2050 the level of income supplement being received by the bottom quintile ($25,710) is greater than the income received through earnings ($24,540).

Appendices 27

SCENARIOS FOR THE 21

ST

CENTURY

APPENDIX A4. ECONOMIC INEQUALITY 1. Definitions and Measures Measures of economic inequality used in this exercise include: a) frequency distributions and histograms. A frequency distribution shows the number of income units (individuals, families, nations, etc.) that fall within a set of income ranges. The graphical display of this data is a histogram. Box IIB-9 in the text shows a simple frequency distribution and histogram of household income in 1992. b) percentile ratios. A simple measure of the inequality of a frequency distribution is the ratio between two percentiles from opposite ends of the distribution. The 80/20 and the 90/10 ratios are commonly used. The 80/20 ratio is calculated by taking the ratio of the total income (or mean income) of all income units in the fifth quintile of a population to that of the first, or bottom, quintile. The 80/20 ratio of the frequency distribution of household income shown in Table 1 of Box IIB-10 is 91.5 / 7.3, or 12.4 to 1. c) Lorenz curves. The Lorenz curve is a graph of cumulative share of a variable, such as income or wealth, against the cumulative population share. The greater the convexity of the Lorenz curve with respect to the bottom right corner of the display, the greater the inequality of the distribution. The diagonal line represents a distribution of perfect equality. The Lorenz curve of the frequency distribution shown in Table 1 of Box IIB-10 is shown in Figure 1 of that box. d) Gini coefficient. This is the most commonly used single statistic of inequality. It is the ratio of the area enclosed by the Lorenz curve and the diagonal, and the total area under the diagonal line. A Gini coefficient of 0.0 indicates perfect equality, and a ratio of 1.0 indicates perfect inequality. Some authors multiply the fractional Gini coefficient by 100, so that 0

Appendices 28

SCENARIOS FOR THE 21

ST

CENTURY

indicates perfect equality and 100 indicates perfect inequality, as a convenience. The Gini coefficient can be calculated directly from a frequency distribution by using the formula2:

where: N= the number of percentiles y = the relative shares of income of each percentile y bar = the mean income share of the percentiles i = the index number of the percentiles (1, 2, N) 2. Interpreting the Gini coefficient In general, one or two Gini points are not very significant, but ten Gini points is very significant. Box A4-1 shows quintile distributions that typify groups of countries situated at different points along the gradient of inequality as measured by the Gini coefficient. We see that the 20 countries with the greatest income equality have an average Gini of about 28, while the 28 with the least income equality have a Gini of about 54. Box A4-2 shows the amount of redistribution that would be required to change the Gini coefficient of the United States by one Gini point, from its current .37 to .36. This could be done by a tax of 5.7% on the incomes of the top quintile (Q5) of households, with the proceeds transferred to the next lowest quintile (Q4). Or it could be done if a tax of 1.5% was levied on Q5 but transferred to the bottom quintile (Q1).

See Cowell (1995) and Lambert (1993) for a review the technical tools for measuring economic inequality.

Appendices 29

SCENARIOS FOR THE 21

ST

CENTURY

BOX A4-1. COMPARISON OF INCOME DISTRIBUTIONS

[source: Box IIB-1]

The ve displays illustrate the sorts of income distributions that characterize different countries. See the key at the bottom. income share (%) Distribution A: bottom 20% second 20% middle 20% fourth 20% top 20% Distribution B: bottom 20% second 20% middle 20% fourth 20% top 20% Distribution C: bottom 20% second 20% middle 20% fourth 20% top 20% Distribution D: bottom 20% second 20% middle 20% fourth 20% top 20% Distribution E: bottom 20% second 20% middle 20% fourth 20% top 20% 8.6 13.2 15.3 27.4 35.5

Income Distribution A 70 70

Income Distribution B

Share of Total Income (%)

50 40 30 20 10 0 bottom 20% second 20% middle 20% fourth 20% top 20%

Share of Total Income (%)

60

60 50 40 30 20 10 0 bottom 20% second 20% middle 20% fourth 20% top 20%

6.3 9.6 15.3 30.2 38.6

5.8 8.7 15.3 30.0 40.2

Share of Total Income (%)

Income Distribution C 70 70

Income Distribution D

5.1 7.1 9.9 34.2 43.7

50 40 30 20 10 0 bottom 20% second 20% middle 20% fourth 20% top 20%

Share of Total Income (%)

60

60 50 40 30 20 10 0 bottom 20% second 20% middle 20% fourth 20% top 20%

3.4 4.0 5.6 30 57

Income Distribution E 70

Share of Total Income (%)

60 50 40 30 20 10 0 bottom 20% second 20% middle 20% fourth 20% top 20%

Income Distribution A B C D E

descriptions 20 most equal countries, ranked by Gini next 20 countries, ranked by Gini next 20 countries, ranked by Gini next 20 countries, ranked by Gini 28 most unequal countries, ranked by Gini

mean Gini 27.6 34.5 35.5 43.5 54.1

mean 80/20 4.1 6.1 6.9 8.5 17.1

Typical Examples

Canada, Russia, Spain, Bangladesh Germany, Indonesia, Ghana, India Japan, USA, China, Nigeria Australia, Peru, Uganda, Iran Mexico, Kenya, Thailand, Brazil

Appendices 30

SCENARIOS FOR THE 21

ST

CENTURY

BOX A4-2. Changes in Income Distribution Needed to Change the Gini Coefcient of the United States by 1 Point Transfers between Quintiles: current % shares ('92) Q5 => Q4 Q5 => Q3 Q5 => Q2 4.5 10.7 16.6 24.1 44.1 0.37 9.8 26.6 41.6 0.36 9.2 17.9 42.8 0.36 9.5 11.6 43.2 0.36 9.6

quintiles Q1 Q2 Q3 Q4 Q5 lowest 20% second 20% middle 20% fourth 20% top 20%

Q5 => Q1 5.15 43.45 0.36 8.4

Gini 80/20

Q1 Q2 Q3 Q4 Q5

lowest 20% second 20% middle 20% fourth 20% top 20%

mean Q5 => Q4 Q5 => Q3 Q5 => Q2 Q5 => Q1 income 5553 6355 13204 14314 20484 22089 29739 32824 54419 51334 52815 53309 53617

Q1 Q2 Q3 Q4 Q5

lowest 20% second 20% middle 20% fourth 20% top 20%

gross tax & Q5 => Q4 Q5 => Q3 Q5 => Q2 Q5 => Q1 transfer (%) 14.4% 8.4% 7.8% 10.4% -5.7% -2.9% -2.0% -1.5%

Q1 Q2 Q3 Q4 Q5

lowest 20% second 20% middle 20% fourth 20% top 20%

amounts transferred -

Q5 => Q4 3085 -3085

Q5 => Q3 1604 -1604

Q5 => Q2 1110 -1110

Q5 => Q1 802 -802

Appendices 31

SCENARIOS FOR THE 21

ST

CENTURY

APPENDIX 5

DEFINITIONS OF GROWTH, WELL-BEING AND RELATED KEY CONCEPTS

As most commonly used, the word growth simply means an increase. In many contexts growth applies more generally to any change over time, and thus allows negative growth. In these notes we might be interested in the growth, or change over time, of output, consumption, utility (or welfare, well-being or happiness), throughput (or resource use), population, pollution, the stock of natural or human-made capital, information, complexity or any of many other things. For our purposes it is particularly important to have clear definitions of well-being, throughput and output. There is a large literature on how these might be defined and measured, but the rough definitions below will serve our purposes for now. Well-being refers to the most general, inclusive sense of human satisfaction and contentment. The question what determines well-being? is among the most important questions there are. Throughput is the flow of natural resources used to generate output. Some throughput generates output directly, but most becomes output only after a series of manipulations. During these manipulations some of the throughput becomes waste. Particular flows of resources can be easily measured but there is no practical common unit with which to measure throughput in the aggregate. Output is throughput manipulated by human activities intended to increase well-being. It can equivalently be defined as all final goods and services produced by an economy during a time period. To the extent that prices reflect value, output can be measured as the sum of the prices of these goods and services, or gross domestic product (GDP). The proper measurement of output, or GDP, should include imputed prices for both positive and negative non-market goods. While any particular act of producing output is intended to increase well-being, it may not. As

Appendices 32

SCENARIOS FOR THE 21

ST

CENTURY

defined here output has an important bearing on well-being but is hardly synonymous with it. In these notes the term economic growth is used to mean the growth of output. Well-being, throughput and output can each increase, decrease or remain constant. We can imagine 3^3 = 27 scenarios, some more plausible than others, showing how these three variables might change with respect to one another over time. Three scenarios are shown in Figures 1, 2 and 3 in A5-1. Figure 1 illustrates a conventional interpretation of the historical experience of the industrial world over the past three centuries or so. Greater throughput enables production of more output, which increases well-being. Figure 2 is a green scenario. Technological innovation allows a constant level of output to be produced with less throughput, while skillful craftwork changes qualities of that output in such a way that well-being increases. Figure 3 is a mixed bag. A constant level of throughput is maintained, perhaps because a transition to a renewables-based economy has been successfully completed. And technological innovation is robust enough to generate a steady increase in output. But well-being declines nonetheless, perhaps because this materially sustainable, technologically productive world is being run as an increasingly authoritarian police state. As defined, output represents a link between the fundamentally more important variables of well-being and throughput. Output is of interest to us primarily because it contributes to wellbeing and uses throughput. The five scenarios shown in Figures 1 through 5 in Box A5-2 dispense with output and illustrate different sorts of relationships that may obtain more directly between the level of wellbeing and the level of throughput.

Appendices 33

SCENARIOS FOR THE 21

ST

CENTURY

BOX A5-1. WHAT GROWS? (1) The figures show well-being, output and throughput increasing, decreasing and remaining constant over time. well-being 1. output throughput

time 2. well-being output throughput

3. . well-being output throughput

Appendices 34

SCENARIOS FOR THE 21

ST

CENTURY

BOX A5-2.

WHAT GROWS? (2)

The figures show different ways that trajectories of throughput might bear upon trajectories of well-being, as discussed in the text. well-being throughput throughput well-being
time

1. the metal eaters

2. gloomy Sunday

well-being

well-being

technological change throughput 3. techno-green

subjective change throughput 4. subjective green

well-being throughput

5. green minimal claim

Appendices 35

SCENARIOS FOR THE 21

ST

CENTURY

Figure 1 illustrates, once more, the conventional account.3 Figure 2 shows the unhappy case in which throughput continues to grow but well-being declines. Figures 3 and 4 are both green scenarios and show the same formal relationship between throughput and well-being, but they are driven by very different processes and represent very different outcomes. Figure 3 shows the techno-green scenario. Human technological skill allows a constant level of throughput to be manipulated in increasingly efficient and innovative ways. As a result, human well-being can increase without endangering the environment. Figure 4 is the subjective-green scenario. It shows a constant level of throughput, but this time there are no technological improvements. The same stuff is made in the same way, century after century. However, the quality of the services provided by the output improves. The design and construction of a saxophone may not change over the course of a century, but the beauty and variety of the compositions written for it, and of the performances given with it, can grow indefinitely. There is a second, more subtle interpretation of the scenario shown in Figure 4. Neither the saxophone, the compositions, nor the performances change over the course of time, but the listeners develop an increasingly keener appreciation for the music. Throughput remains constant but well-being grows. Figure 5 is the green minimal claim scenario. The minimal claim is simply that a constant level of throughput can indefinitely support some constant, fulfilling level of well-being. Limits and constraints on the rates and levels of growth The phrase limits to growth is ambiguous because the thing whose growth is limited is unspecified. It is also ambiguous because growth can refer either to the growth rate or to the level to which something can grow. We can say that output (for example) is limited to a growth

The metal eaters is a translation of , used by Russian environmentalists to describe the institutions of industrial civilization.

Appendices 36

SCENARIOS FOR THE 21

ST

CENTURY

rate of 2% per year, or that output is limited to an absolute level of $30 trillion. Note that a limit to the level of growth implies an eventual limit to the rate of growth, i.e., 0%. The term limits can be used to mean either an absolute level that cannot be exceeded, or a softer set of constraints short of an absolute limit. In this exercise we try to use limits when speaking of values that cannot be exceeded, and constraints for the more general set of impediments. As used here, constraints is the inclusive term: a limit is a very strong constraint, but a constraint need not, although it might, impose a limit. Thus we can speak, hopefully unambiguously, of limits to the level to which something can grow, limits to the rate at which something can growth, constraints on the level to which something can grow, and constraints on the rate at which something can grow. The Rhetoric of Growth President Ronald Reagan once offered the opinion that there are no limits to growth, because there are no limits to the human imagination. Put less succinctly but more precisely, he might have been expressing the opinion that there are no limits to the ability of technology to manipulate any given level of throughput such that human well-being continues to increase. Some environmentalists say, There are limits to growth, but not to well-being. One interpretation of this statement might be that technology will allow us to manipulate an environmentally sustainable, constant level of throughput in a way that can allow human wellbeing to increase indefinitely. Other environmentalists say, There are limits to throughput, but there are no limits to growth. In the previous statement growth referred to throughput, but here it could refer to well-being, or to output, or to both well-being and output. Can We Measure Well-Being? The Standard National Accounts were developed in the 1940s in order to track the flows of economic variables, including output, consumption, saving and public spending, that were needed to help develop policies intended to ensure steady output growth and low unemployment.

Appendices 37

SCENARIOS FOR THE 21

ST

CENTURY

In the wake of the Great Depression and the Second World War, a steady growth in output was experienced by most persons as a major contribution to well-being. The availability of GDP as an analytically convenient measure of output led to its identification among professionals and the public as a measure of well-being. The inadequacies of conventional GDP as a measure of both output and well-being are well known. Its major inadequacy as a measure of output is that it doesnt account for goods or bads external to the market. Attempts to deal with the shortcomings of the GDP have gone in two directions. One is to strengthen its use as a measure of output, while disclaiming any role for it as a measure of well-being. This entails converting as many non-market goods and bads as possible into dollar terms and incorporating these into the national income accounts, and other reforms. A second direction seeks to devise genuine measures of well-being which over time could supplant GDP in that role. One approach involves the use of (fully inclusive) GDP in association with satellite accounts of variables that are judged to have a bearing on well-being, such as literacy rate, longevity, infant mortality, or stocks and flows of natural resources. A second approach is to combine these measures into a single statistic by means of an indexing formula. The choice set of satellite accounts, and any indexing weights they might be given, are of course strongly normative statements. Some examples of efforts along the lines described are shown in A5-3. Figure 1 shows the measure that Nordhaus and Tobin (1989) call Net Economic Welfare (NEW). NEW takes Net National Product, which is GNP less depreciation, and adds the value of leisure time, domestic activities, underground transactions, and other non-market goods. It subtracts nonmarket bads such as economic damages resulting from pollution and urban congestion. Nordhaus defines NEW as an adjusted measure of total national output that includes only consumption and investment items that contribute directly to economic well-being. In general, the growth rate of NEW has paralleled the growth rate of conventional GDP.

Appendices 38

SCENARIOS FOR THE 21

ST

CENTURY

BOX A5-3. MEASURING WELL-BEING These figures illustrate how differing measures of well-being have changed in recent years. Figure 1. Net Economic Welfare
(Reprinted from Samuelson and Nordhaus, 1989, p 119)

Figure 2. GPI versus GDP

(Reprinted from Redefining Progress, 1995)

Appendices 39

SCENARIOS FOR THE 21

ST

CENTURY

Box A5-3. MEASURING WELL-BEING (cont.) Figure 3. Human Development Index, 1960-1993
(United Nations Development Program, Human Development Report, 1996)

Figure 4. Happiness (Reprinted from Easterlin, 1995)

The vertical axis shows the percent very happy, based on National Opinion Research Center for Genetics and Society surveys through 1991.

Appendices 40

SCENARIOS FOR THE 21

ST

CENTURY

Figure 2 displays the Genuine Progress Indicator (GPI) developed by the organization Redefining Progress (1995), based on work by Daly and Cobb (1989). Like NEW, the GPI adds the value of domestic activities to GDP. However, the GPI shows a decrease in leisure time rather than an increase. In the GPI environmental pollution and resource depletion reduce GDP by larger amounts than they do in the NEW. In addition, GPI incorporates an indexing factor that records the growth of economic inequality as a decrease in well-being. These plus other modifications of the GDP accounts generate a path over time very different from that displayed by NEW. Figure 3 shows the Human Development Index developed by the United Nations Development Program (1996). This relatively simple index has three components: life expectancy, educational attainment (based on adult literacy rates and mean years of schooling), and adjusted income. Income is GDP per capita computed on a Purchasing Power Parity basis, subject to a strongly diminishing marginal index value. Easterlin (1995) doubts that any set of objective measures can accurately model all the determinants of well-being, and suggests that surveys of how people assess their own well-being can provide a better guide. Figure 4 shows how people in the United States responded to a question asking them how happy they are. The slight trend shown by the regression line is not statistically significant. Reported happiness in the United States over the past twenty years has been constant, despite a major increase in output during that time. Applied to the United States, these alternatives to conventional GDP move in different ways over time. The NEW and HDI increase, the GPI decreases, and the proportion of people stating that they are happy remains unchanged.

Appendices 41

SCENARIOS FOR THE 21

ST

CENTURY

APPENDIX 6. ESTIMATES OF STOCKS OF ENERGY RESOURCES Boxes A6-1 through A6-7, below, show the source data used to estimate stocks of energy resources used in Section IIA.2.a. Box A6-8 shows the conversion factors used to convert all estimates to terawatt years. Box A6-1. HOLDREN source: Energy & Resources 200 course handout: Recent World Energy and Economic Data, 10/95. Estimated Remaining Recoverable Resources of Fuels (TWy) oil petroleum liquids oil shale heavy oils natural gas conventional unconventional coal TOTAL Fossil Fuels uranium (in LWRs) uranium (in LMFBRs) lithium (DT fusion) deuterium (fusion) 500 30000 500 500 1000 5000 37500 3000 3000000 1.4E+08 2.5E+11 7500 (w/o oil shale)

Box A6-2. World Energy Council (source: Global Energy Perspective to 2050 and Beyond. Report, 1995. p 36.) Estimates of Energy Reserves, Resources and Occurrences (GTOE) Reserves Resources Resource Additional ultimately Base Occurrencesrecoverable oil conventional 150 145 295 200 unconventional 193 332 525 1900 553 (511-595) natural gas conventional 141 279 420 220 unconventional 192 258 450 400 hydrates 18700 coal 606 2794 3400 3000 3400 TOTAL Fossil Fuel: uranium (LWRs) uranium (LMFBRs) TOTAL ALL: 1282 57 3390 4729 3808 203 12150 16161 5090 260 15540 20890 24000 150 8900 33050

World Energy Council: conversion to terawatt years Estimates of Energy Reserves, Resources and Occurrences (TWy) 1GTOE = 1.42 TWy Reserves oil conventional unconventional natural gas conventional unconventional hydrates coal TOTAL Fossil Fuels uranium (LWRs) uranium (LMFBRs) TOTAL ALL: Resources Resource Base 206 471 396 366 3967 5407 288 17253 22949 419 746 596 639 4828 7228 369 22067 29664 Additional ultimately Occurrencesrecoverable 284 785 312 568 26554 4260 34080 213 12638 46931 (+ 17,000 tons U ultimately recoverable)

213 274 200 273 861 1820 81 4814 6715

2698

4828

Appendices 42

SCENARIOS FOR THE 21

ST

CENTURY

Box A6-3. IPCC (source: IPCC, Climate Change 1995: Impacts, Adaptations and Mitigations of Climate Change. Working Group II Report) (EJ) Reserves Identied oil conventional unconventional natural gas conventional unconventional hydrates coal TOTAL Fossil Fuels uranium (LWRs) TOTAL ALL: 6000 7100 4800 6900 25200 50000 1800 51800 4500 4500 6900 2300 9200 16400 16400 Conventional Resources Unconventional Resources: Resource to be discovered. w/ probability: a. currently b. recoverable with Base 0.95 0.5 0.05 recoverable technological progress 1800 2500 5500 9000 2700 4400 10900 2200 13900 16100 4100 20200 17800 86400 113200 6000 119200 8500 16100 9200 26900 125500 186200 14200 200400 Additional Occurance ( > than...) 10000

10000 22000 800000 130000 987000 1000000 1987000

IPCC: conversion to terawatt years (TWy) 1 EJ = 0.03175 TWy Reserves Identied oil conventional unconventional natural gas conventional unconventional hydrates coal TOTAL Fossil Fuels uranium (LWRs) TOTAL ALL: 191 225 152 219 800 1588 57 1645 143 143 219 73 292 521 521 Conventional Resources Unconventional Resources: Resource to be discovered. w/ probability: a. currently b. recoverable with Base 0.95 0.5 0.05 recoverable technological progress 57 79 175 286 86 140 346 70 441 511 130 641 565 2743 3594 191 3785 270 511 292 854 3985 5912 451 6363 Additional Occurance

318

318 699 25400 4128 31337 31750 63087

Box A6-4. ENERGY MODELING FORUM (source: Demographic, Economic, and Energy Assumption, EMF 14. http://soe.stanford.edu/ees/design.html) (EJ) Reserves crude oil + nat. gas liquids natural gas coal Undiscovered Resources, 95%-ile 6833 10805 300000 ultimately recoverable resources total

7362 5537

14195 16342 300000

ENERGY MODELING FORUM: conversion to terawatt years (source: Demographic, Economic, and Energy Assumption, EMF 14. http://soe.stanford.edu/ees/design.html) (TWy) 1 EJ = 0.03175 TWy Reserves Undiscovered ultimately recoverable total Resources, 95%-ile resources crude oil + nat. gas liquids 234 217 451 natural gas 176 343 519 coal 9525 9525

Appendices 43

SCENARIOS FOR THE 21

ST

CENTURY

Box A6-5 . HERMAN KAHN: THE NEXT 200 YEARS (1976) (10^18 Btu) source oil natural gas coal shale oil tar sands total fossil fuels U-235 (free world) U-235 (oceans) Uranium for breeders Li-6 (D-T fusion) Deuterium (D-D fusion) proven long term reserves potential resources 3.7 14.4 1 15.8 95 170 19 2000 1.8 1.8 120.5 2202 15 3000 100000 + 320 1E+09 +

HERMAN KAHN: THE NEXT 200 YEARS (1976) Conversion toTerawatt years 10^18 Btu = 33.49 TWy source oil natural gas coal shale oil tar sands total fossil fuels U-235 (free world) U-235 (oceans) Uranium for breeders Li-6 (D-T fusion) Deuterium (D-D fusion) proven long term reserves potential resources 124 482 33 529 3182 5693 636 66980 60 60 4036 73745 502 100470 3349000 + 10717 33490000000 +

BOX A6-6. ROGER HINRICHS: Energy: Its Use and Environment (text; 1996) 10^18 Btu proven reserves oil natural gas coal tar sands shale oil total fossil fuel 5.9 5 50 1.7 0.87 63.47

ROGER HINRICHS: Energy: Its Use and Environment (text; 1996) Conversion to Terawatt years 10^18 Btu = 33.49 TWy proven reserves oil natural gas coal tar sands shale oil total fossil fuel 198 167 1675 57 29 2126

Appendices 44

SCENARIOS FOR THE 21

ST

CENTURY

BOX A6-7. Freeman & Jahoda: WORLD FUTURES, 1978 (10^9 tce) proven & ultimately fuel possible reserves recoverable resources coal oil natural gas shale oil tar sand sub-total, for fossil hydrocarbons Uranium thermal Uranium breeder Thorium breeder Fusion geothermal wet rock geothermal dry rock TOTAL (excl fusion & geothermal dry rock) 1000 375 200 7700 2760 1500 1200 225

1575 100 0 0 0 60 0

13385 ? 5000 + 4000 + virtually unlimited 1500 50 million

1735

23885

Freeman & Jahoda: WORLD FUTURES, 1978 Conversion to Terawatt years (10^9 tce) 10^9 tce = 0.929 TWy proven & ultimately fuel possible reserves recoverable resources coal oil natural gas shale oil tar sand sub-total, for fossil hydrocarbons Uranium thermal Uranium breeder Thorium breeder Fusion geothermal wet rock geothermal dry rock TOTAL (excl fusion & geothermal dry rock) 929 348 186 7153 2564 1394 1115 209

1463 93 0 0 0 56 0

12435 ? 4645 + 3716 + virtually unlimited 1394 50 million

1612

22189

BOX A6-8.

CONVERSION FACTORS

The conversion factors used in Boxes A6-1 through A6-7 are: 1 EJ = 10^18 J 1 TWy = 31.5 EJ 1 EJ = 3.175 x 10^-2 TWy 1 J = 9.484 x 10^-4 Btu 1 Btu = 1055 J 1 Q = 10^18 Btu 1 bbl = 5.8 x 10^6 Btu = 6.12 x 10^9 J 1 ton U235 = 70 x 10^12 Btu 1 mbdoe = 50 x 10^6 tons oil equivalent/year = 2.2 x 10^18 J/yr 10^15 Btu = 36 x 10^6 tons coal equivalent where: EJ = Exajoule TWy = Terawatt year Btu = British thermal unit Q = Quad bbl = billion barrels of oil equivalent mbdoe = million barrels per day oil equivalent

Appendices 45

SCENARIOS FOR THE 21

ST

CENTURY

APPENDIX 7. Calculation of Global Warming Trajectories Section II.A.2a offered several scenarios in which anthropogenic global warming is prevented from exceeding 3.6o C, as displayed in Box IIA-15. These scenarios were generated using my extension (Hayes 1995) of the model developed by Cline (1992). The key equations of the extended Cline model are shown and described below. We begin with Clines equations for income and population:

where: y = per capita income P = population D = developed countries L = less developed countries t = time; t = 1 in 1991 g = per capita growth in income n = per capita growth in population These give Gross World Product, Y:

Clines proposed carbon emissions ceiling (K) [4 Gt/year] is subtracted from baseline emissions (E) to give the needed amount of annual emissions reductions (R):

The amount of global warming (W) in any given year (t) is given by the linear functions:

Appendices 46

SCENARIOS FOR THE 21

ST

CENTURY

where: S = climate sensitivity parameter [1.5, 2.5, and 4.5 for low, medium and high cases] W* = equilibrium warming under very-long-term global warming conditions [6, 10 and 18 for low, central and high climate sensitivity parameters] I extended Clines model to allow alternative warming trajectories to be calculated. The equations are shown below. Further details can be found in Hayes (1995). First apply the subscript t to K, to show that K can vary with time, with t0 = 1966.

Next we add the climate model, beginning with equations for atmospheric CO2 stock and concentration. For this purpose I used an equation from the Nordhaus (1996) DICE model. The superscript denotes business-as-usual values.

In this equation 590 is the pre-industrial carbon stock (in GtC), is the short-term emissions retention fraction (.64), Et are emissions, and is the long-term CO2 decay constant, set at .00825.

Next we multiply by the constant .47 to convert atmospheric stock to a concentration:

Next we use the standard formula to calculate radiative forcing due to CO2:

Then we add the radiative forcing that results from trace gases:

Now we calculate warming commitment:

Appendices 47

SCENARIOS FOR THE 21

ST

CENTURY

In this equation is .3 w/m^2 and the feedback variable takes upper bound, central case and lower bound values of 3.4, 1.9, and 1.1, respectively, which generate climate sensitivity factors of 1.5, 2.5 and 4.5 degrees C. The factor , set at 1.18, adjusts for the masking effects of SO2, NOx and O3, and for the use of Nordhaus' stock equation. Finally we get warming at any time t by adjusting for the 25-year-lag:

Appendices 48